Files in this item

FilesDescriptionFormat

application/pdf

application/pdfLIU-DISSERTATION-2017.pdf (3MB)
(no description provided)PDF

Description

Title:Decentralized optimization approach for power distribution network and microgrid controls
Author(s):Liu, Hao Jan
Director of Research:Zhu, Hao
Doctoral Committee Chair(s):Zhu, Hao
Doctoral Committee Member(s):Dominguez-Garcia, Alejandro; Sauer, Peter; Srikant, Rayadurgam
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:Ph.D.
Genre:Dissertation
Subject(s):Voltage control
Volt/volt-ampere reactive (VAR) control
Frequency control
Distribution networks
Microgrids
Cybersecurity
Optimization
Abstract:The smart grid vision has led to the development of advanced control and management frameworks using distributed generation (DG) and storage resources, commonly referred to together as distributed energy resources (DERs). Albeit environment-friendly, these DERs in distribution networks including microgrids (MGs) could greatly challenge the operational goal of maintaining adequate power system reliability standards because of their high intermittency, uncertainty, and lack of physical inertia. Meanwhile, these networks are inherently unbalanced and lack high-quality communications to a centralized entity as compared to the bulk transmission grid. Both aspects contribute to the challenge of designing voltage and frequency control frameworks therein. To tackle these problems, we propose decentralized control strategies, which account for cyber-physical network interactions automatically and dynamically while being either cognizant of various communication scenarios or resilient to malicious cyber intrusions. By treating the transmission grid as an infinity bus, voltage stability is the main concern in distribution networks where more DERs are being installed in the near future. Thanks to advances in power electronics, DERs can also be excellent sources of reactive power (VAR), a quantity that is known to have a significant impact on the network voltage level. Accordingly, we first formulate the local VAR-based voltage control design by minimizing a weighted quadratic voltage mismatch error objective using gradient-projection (GP) updates. The step-size design under both static and dynamic settings is further analyzed for practical implementation purposes. Nonetheless, such local design suffers degraded performance due to lack of information exchanges, especially under limited VAR resources. To address this issue, we develop the distributed voltage control (DVC) design based on the alternating direction method of multipliers (ADMM) algorithm. The DVC design has simple node-to-node communication architecture while seamlessly adapting to dynamically varying system operating conditions and being robust against random communication link failures. To further reduce communication complexity and enhance robustness to imperfect communications, especially under the worst-case scenarios of a total communication outage, we integrate both local and distributed control designs to a hybrid voltage control (HVC) scheme that can achieve the dual objectives in terms of flexible adaptivity to variable rate of communications and global optimality of voltage regulation performance. Such an innovative design aims to unify the separated framework of either local or distributed control design. Numerical tests using realistic feeders and real time-series data have been demonstrated for the voltage control designs. The aforementioned decentralized voltage control designs can improve the power system stability while distribution feeders are interconnecting to the bulk transmission grids. With a high penetration of DERs in the networks, it is possible to build a discrete energy system, namely, a microgrid (MG), that is capable of operating in parallel with, or independently from, the transmission grids. Henceforth, MGs are likely to emerge as a means to advance power and cyber physical resiliency in future grid systems. As MGs may operate independently, these mostly power electronics-interfaced DERs exhibiting low-inertia characteristic have raised significant concern over the frequency stability issues. To tackle this problem, we introduce the concept of virtual inertia of DERs and cast the secondary frequency control design for isolated MGs as a consensus optimization problem. We solve it distributively by adopting the partial primal-dual (PPD) algorithm. Interestingly, parts of our specially designed control algorithm turn out to mimic the dynamics of network power flow and virtual synchronous generator-based inverter. Thus, such dynamics is seamlessly governed by the physical system itself. Given a proper control parameter choice, the convergence of the consensus is guaranteed without assuming the time-scale separation of the hierarchical control design methodologies. By extending this work to a practical industrial MG network that follows the IEC 61850 communication protocol, similar frequency regulation objective is introduced and solved by a decentralized ADMM-based algorithm. The countermeasures for malicious attacks on the communication network for both PPD- and ADMM-based control designs are also investigated. Specifically, we analyze two types of malicious attacks on the communication network, namely, the link and node attacks. Meanwhile, anomaly detection and localization strategies are developed based on the metrics of optimization-related variables. We showcase the microgrid frequency regulation operation to demonstrate the effectiveness of the proposed frequency control designs under a real-time simulation environment.
Issue Date:2017-12-07
Type:Text
URI:http://hdl.handle.net/2142/99393
Rights Information:Copyright 2017 Hao Jan Liu
Date Available in IDEALS:2018-03-13
Date Deposited:2017-12


This item appears in the following Collection(s)

Item Statistics